Range and/or polarity-based thresholding for improved data extraction

• US 10,467,465 B2
• Filed: 12/30/2016
• Issued: 11/05/2019
• Est. Priority Date: 07/20/2015
• Status: Active Grant

• Current Assignee: Tungsten Automation Corp.
• Original Assignee: Kofax Incorporated
• Inventors: Shustorovich, Alexander, Thrasher, Christopher W.
• Primary Examiner(s): Motsinger, Sean T
• Application Number:
  US15/396,327
• Publication Number:
  US 20170109576A1
• Time in Patent Office: 1,040 Days
• CPC Class Codes: G06T 2207/20104 Interactive definition of region of interest [ROI]
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: G06T 7/187 involving region growing; involving region merging; involving connected component labelling
• CPC Class Codes: G06V 10/82 using neural networks
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/15 Cutting or merging image elements, e.g. region growing, watershed or clustering-based techniques
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/162 Quantising the image signal
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

Computerized techniques for improved binarization and extraction of information from digital image data are disclosed in accordance with various embodiments. The inventive concepts include: rendering, using a processor of the mobile device, a digital image using a plurality of binarization thresholds to generate a plurality of range-binarized digital images, wherein each rendering of the digital image is generated using a different combination of the plurality of binarization thresholds; identifying, using the processor of the mobile device, one or more range connected components within the plurality of range-binarized digital images; and identifying, using the processor of the mobile device, a plurality of text regions within the digital image based on some or all of the range connected components. Corresponding systems and computer program products are also disclosed.

Claims

1. A computer program product comprising a non-transitory computer readable storage medium having embodied thereon computer readable program instructions configured to cause a processor, upon execution of the computer readable program instructions, to perform operations comprising:

rendering, using the processor, a digital image using a plurality of binarization thresholds to generate a plurality of range-binarized digital images, wherein each rendering of the digital image is generated using a different combination of the plurality of binarization thresholds, and wherein each combination of the plurality of thresholds includes a unique upper threshold and a unique lower threshold;

identifying, using the processor, one or more range connected components within the plurality of range-binarized digital images; and

identifying, using the processor, a plurality of text regions within the digital image based on some or all of the range connected components.

2. The computer program product as recited in claim 1, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to:

identify, using the processor, one or more connected components of black pixels within the plurality of binarized digital images; and

wherein the plurality of text regions are identified further based on some or all of the connected components of black pixels identified within the plurality of binarized digital images.

3. The computer program product as recited in claim 1, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to:

identify, using the processor, one or more connected components of white pixels within the plurality of binarized digital images; and

wherein the plurality of text regions are identified further based on some or all of the connected components of white pixels identified within the plurality of binarized digital images.

4. The computer program product as recited in claim 1, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to evaluate, using the processor, a confidence of an identity of each of the range connected components identified within the plurality of range-binarized digital images;

and

wherein the one or more range connected components within the plurality of range-binarized digital images that are utilized to identify the plurality of text regions within the digital image consist of range connected components having an identity confidence level value above a predefined identity confidence threshold.

5. The computer program product as recited in claim 1, wherein identifying the plurality of text regions within the digital image comprises:

calculating one or more geometric characteristics of the one or more range connected components identified within the plurality of range-binarized digital images;

grouping adjacent of the one or more connected components that exhibit one or more common geometric characteristics;

defining each grouping of three or more adjacent range connected components that exhibit the one or more common geometric characteristics as a candidate text region; and

assembling the candidate text regions by removing overlaps to identify the plurality of text regions within the digital image.

6. The computer program product as recited in claim 5, wherein the one or more geometric characteristics are selected from:

a connected component baseline, a connected component top line, a connected component height, a connected component stroke thickness;

a connected component density;

a connected component number of vertical and horizontal crossings; and

color information corresponding to the connected component, the color information being represented in or derived from the digital image.

7. The computer program product as recited in claim 1, wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a darker text character represented on a lighter background within the digital image.

8. The computer program product as recited in claim 1, wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a lighter text character represented on a darker background within the digital image.

9. The computer program product as recited in claim 1, wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a text character represented on a dual background within the digital image, at least some of the dual background being lighter than the text character, and at least some of the dual background being darker than the text character.

19. The computer program product as recited in claim 1, wherein each combination of the plurality of thresholds encompasses a range of intensity value units from about 5 units to about 10 units.

10. A computer-implemented method, comprising:

rendering, using a processor, a digital image using a plurality of binarization thresholds to generate a plurality of range-binarized digital images, wherein each rendering of the digital image is generated using a different combination of the plurality of binarization thresholds;

identifying, using the processor, one or more range connected components within the plurality of range-binarized digital images;

identifying, using the processor, a plurality of text regions within the digital image based on some or all of the range connected components; and

wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a text character represented on a dual background within the digital image, at least some of the dual background being lighter than the text character, and at least some of the dual background being darker than the text character.

11. The method as recited in claim 10, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to:

identify, using the processor, one or more connected components of black pixels within the plurality of binarized digital images; and

wherein the plurality of text regions are identified further based on some or all of the connected components of black pixels identified within the plurality of binarized digital images.

12. The method as recited in claim 10, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to:

identify, using the processor, one or more connected components of white pixels within the plurality of binarized digital images; and

wherein the plurality of text regions are identified further based on some or all of the connected components of white pixels identified within the plurality of binarized digital images.

13. The method as recited in claim 10, comprising computer readable program instructions configured to cause the processor, upon execution of the computer readable program instructions, to evaluate, using the processor, a confidence of an identity of each of the range connected components identified within the plurality of range-binarized digital images;

and

wherein the one or more range connected components within the plurality of range-binarized digital images that are utilized to identify the plurality of text regions within the digital image consist of range connected components having an identity confidence level value above a predefined identity confidence threshold.

14. The method as recited in claim 10, wherein identifying the plurality of text regions within the digital image comprises:

calculating one or more geometric characteristics of the one or more range connected components identified within the plurality of range-binarized digital images;

grouping adjacent of the one or more connected components that exhibit one or more common geometric characteristics;

defining each grouping of three or more adjacent range connected components that exhibit the one or more common geometric characteristics as a candidate text region; and

assembling the candidate text regions by removing overlaps to identify the plurality of text regions within the digital image.

15. The method as recited in claim 14, wherein the one or more geometric characteristics are selected from:

a connected component baseline, a connected component top line, a connected component height, a connected component stroke thickness;

a connected component density;

a connected component number of vertical and horizontal crossings; and

color information corresponding to the connected component, the color information being represented in or derived from the digital image.

16. The method as recited in claim 10, wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a darker text character represented on a lighter background within the digital image.

17. The method as recited in claim 10, wherein at least one of the range connected components identified within the plurality of range-binarized digital images corresponds to a lighter text character represented on a darker background within the digital image.

18. The method as recited in claim 10, further comprising normalizing intensity values across one or more color channels of the digital image.

20. A computer-implemented method, comprising:

rendering, using a processor, a digital image using a plurality of binarization thresholds to generate a plurality of range-binarized digital images, wherein each rendering of the digital image is generated using a different combination of the plurality of binarization thresholds;

identifying, using the processor, one or more range connected components within the plurality of range-binarized digital images;

identifying, using the processor, a plurality of text regions within the digital image based on some or all of the range connected components, wherein identifying the plurality of text regions within the digital image comprises;

calculating one or more geometric characteristics of the one or more range connected components identified within the plurality of range-binarized digital images;

grouping adjacent of the one or more connected components that exhibit one or more common geometric characteristics;

defining each grouping of three or more adjacent range connected components that exhibit the one or more common geometric characteristics as a candidate text region; and

assembling the candidate text regions by removing overlaps to identify the plurality of text regions within the digital image.